Malignancies are often characterized by defects in programmed cell death (PCD) pathways contributing to blocks in responses to irradiation and chemotherapy. These defects are frequently manifested by imbalances in the Bcl-2 superfamily of proteins that link survival and death signals to the core PCD machinery. Bcl-2 is over-expressed in about 50% of all cancers. Most chronic lymphocytic leukemias (CLLs) and many Acute Myelogenous Leukemias (AMLs) and Acute Lymphocytic Leukemias (ALLs) over-express anti-apoptotic Bcl-2. Functional studies in vitro suggest an important role for Bcl-2 family proteins in maintaining the survival of these leukemic cells and promoting their resistance to chemotherapy. We hypothesize that apoptosis is controlled by the heterodimerization of competing Bcl-2 family inhibitors, inducers and effectors which ultimately determine whether the inducers channel apoptotic proteins through mitochondrial membranes. Heterodimerization occurs via BH3-domain binding pockets. Preliminary experiments from our laboratory reproducibly demonstrates that a constrained alpha-helical Bak BH3-domain peptide (16 mer) from the pro-apoptotic protein Bak, but not a wild-type unconstrained peptide, overrides block(s) to apoptosis in freshly isolated leukemia cells. The alpha-helical structure is essential for high affinity binding of BH3 peptides, and therefore constrained BH3 peptides are more potent than unconstrained linear peptides. We propose to (1) test various strategems for improving the activity of the Bak BH3 peptide as well as for synthesizing constrained helical BH3 peptides from additional effectors (Bax, Bak) and an inducer (Bid), (2) assess their pro-apoptotic activities and the sensitivities of CLL and AML cells from untreated and relapsed/refractory individuals, (3) identify the targets of pro-apoptotic BH3 peptides by comparing affinities for Bcl-2 family proteins and (4) link a potent pro-apoptotic peptide to membrane permeable peptides for tests against leukemic cells. BH3 peptides could provide powerful tools for proof of concept data in support of efforts to generate small-molecule compounds that mimic Bcl-2 family proteins for the treatment of leukemia and for identifying mechanisms of cell survival and resistance to chemotherapy.